Saturday, November 29, 2008

California Rapid Rail Dreamin'

NOTE: We've moved! Visit us at the California High Speed Rail Blog.

One of the wonderful things about walking down memory lane is that you stumble upon nuggets showing how California has come full circle. Before WW2, Los Angeles and Orange county already had a well-developed network of Red Car trolleys, forerunners of modern light rail.

There is a persistent myth that GM, Standard Oil and Firestone conspired to dismantle the Red Car lines so transit companies would be forced to buy diesel buses instead. The more prosaic truth is that the Pacific Electric company was a loss leader for a real estate tycoon, Henry Huntington. Having made his fortune, he let the quality of service go downhill. The war effort brought rapid growth to Southern California and, the new arrivals had no choice but to live in areas not served by Red Car lines. Planners quickly came to rely on roads and the internal combustion engine to provide residents with the mobility they needed and craved. What had been a rail culture turned into a car culture.

So why am I talking about this on a High Speed Rail blog?

Two reasons:

  1. high speed rail is to short-haul flights as trolleys were to buses (or modern light rail is to bus rapid transit, h/t to DoDo). The pros and cons are comparable, albeit at different speeds and distances. There is no doubt that flying is the more economical choice for truly long distances, in terms of both fare price and the opportunity cost of time spent in transit. Only hard core railfans and pteromechanophobes wax lyrical about spending several days in the comfort of an Amtrak train trundling across the country. At truly short distances in rush hour traffic, subways and light rail reign supreme. It is the middle ground of distances from 30 to roughly 500 miles that HSR will contest in California - with every chance of gaining significant market share.
  2. California voters have decided that the pendulum has swung too far in favor of oil-based mobility and want to redress the balance. The recent rapid run-up in the price of oil triggered the collapse of the housing bubble, forcing US taxpayers to take on over $1,000,000,000,000 in new debt. Add to that the cost of the war in Iraq, which at least indirectly was about retaining access to oil: ka-ching another $500,000,000,000 or more including long-term care for veterans, all on future generations' credit card. That's a whole lot of zeros. The experience has brought the risks inherent in relying on a single source of primary energy, crude oil, into sharp relief. Concerns about anthropogenic climate change come on top of the already known costs of an excessive dependence on crude oil.

Reducing the need to travel and switching to electric vehicles is now in vogue. In the US, most efforts to date have focused on various degrees of electrification of the automobile, but the required batteries must meet multiple conflicting design objectives: high capacity, high power over a wide range of ambient temperatures, life expectancy same as vehicle, crash safety and affordable price tag. You can have any 5 of these 6. Other developed countries have long focused on electrification of their rail networks instead - the technology is ready and proven. That infrastructure is also expensive but funding is collective rather than individual. With proper maintenance, overhead catenaries will last for many decades.

Bottom line: at intermediate distances, electric high speed rail delivers fast service, capacity high enough to avoid expensive airport and freeway expansions, zero tailpipe emissions, excellent safety and, reduced dependence on crude oil in favor of electricity - all at affordable fare prices. CHSRA has gone one further and committed to running all of its trains on renewable electricity.

In the specific case of the US, there are two major stumbling blocks. First, rail is now used primarily for slow, cheap but profitable heavy freight at interstate distances. The infrastructure is owned by competing freight companies, with some trade in trackage rights. Only a few sections of the US rail grid are publicly owned, e.g. the North East Corridor, the Caltrain SF peninsula corridor and, the Alameda freight corridor in LA. Passenger rail volume is very modest by international standards and, taxpayers have long resisted investing in something that only works well if it is perceived as a public service, rather than as a commercial enterprise.

Second, the Federal Rail Administration (FRA) has defined rules that minimize overheads for heavy freight operators. Rather than spend a lot on maintenance, they just keep running a little more slowly every year. They stick with very old, winding alignments with little or no track superelevation to avoid the massive investments needed to support higher speeds. They avoid upgrades to signaling, preferring instead to rely on passive crash safety at low speeds. They do not voluntarily invest in grade separations, diesel exhaust gas aftertreatment or electrification. They even convert dual back to single track alignments and abandon underutilized rights of way altogether, because this reduces their property tax bill. They don't stick to a timetable. The reason for all this is simple: bulk freight customers won't pay a premium for higher speed. All of this works against the provision of effective passenger rail services, as well as regional rail freight of high value goods.

Specifically, the most onerous rule relates to mixed traffic, which in FRA lingo means letting off-the-shelf lightweight rolling stock designed to UIC standards (a.k.a. non-compliant) share track with much heavier equipment that is FRA compliant. By default, this is totally prohibited. Officially, FRA claims its rules improve rail safety. Caltrain has recently shown that the non-compliant electric multiple unit (EMU) rolling stock it would like to migrate to performs as well or better than FRA-compliant equipment in crashes involving road vehicles at grade crossings. The unofficial reason for the mixed traffic rule is that freight operators do not want to invest in the signaling and emergency train control upgrades needed to avoid train-on-train accidents in busy corridors.

One way to address this is to build new rail networks dedicated to rail services other than heavy freight. Typically, that means passenger-only solutions. Examples include subways, light rail, BART and now California HSR. Such systems are allowed to operate non-compliant equipment under a waiver because there is no risk of train-on-train collisions with 15,000 ton heavy freight trains. However, as the recent tragic crash in Chatsworth proved yet again, no amount of crash buffer zones can protect drivers and passengers at even moderate relative speeds. As a result, HR 2095 now mandates the installation of positive train control (PTC) on many corridors on the US national grid by 2015.

Europe and Japan take a different approach to rail safety, one that relies on substantial public investments in track, signaling, rolling stock, maintenance and operator training to avoid accidents. Passive safety is no more than a complement to fall back on if all else fails. The reasoning behind this philosophy is that they consider all rail service, passenger as well as freight, to be to some extent a public service. The intent is to keep people from dying on the roads, to avoid massive investments in road infrastructure and, to reduce exposure to crude oil - almost all of which these countries need to import. Air quality concerns play a subordinate role, the primary focus in on passenger service.

Fortunately, the geography of Europe and Japan lends itself to hauling bulk freight by sea or inland waterways, so rail freight is focused on competing against trucking for relatively fast and punctual delivery of high-value goods, e.g. livestock, foodstuffs, mail/parcels, car parts and cars. In addition, there is the objective of getting trucks off the roads, e.g. in the Alps. Switzerland, Austria/Italy and France/Italy are each working on very long base tunnels to speed freight rail to make it more competitive. Other tunnel projects, such as the Seikan in Japan and the Channel Tunnel in Europe, are focused primarily on replacing slow passenger ferries.

While California HSR is a very worthwhile project in its own right, it also recognizes that passengers rail only works when its stations are in downtown locations with connecting transit and/or within comfortable walking distance of destinations such as large agglomerations of offices, sports stadiums, mega-malls or high-density residential districts. Unfortunately, seismic building codes and the predominance of the automobile have favored low-density development in most areas of the state. In addition, true HSR (top speeds of 186mph or more) can only serve a very limited number of stations because it takes quite a while to reach high speed. Commuter rail and regular-speed intercity trains are supposed to act as feeders into the HSR network. Roughly 10% of proposition 1A funds are reserved for capital investments in existing providers of such services.

However, in order to really drive passenger rail ridership, a bolder approach is needed. If you disregard the very sparsely populated regions north of Sacramento, the Sierras, the Mojave desert and the central coast between Salinas and San Luis Obispo, California is actually quite comparable to many European countries in terms of population density. That means their rail network model and associated philosophy may well make a lot of sense within this one state (with the possible inclusion of the city of Las Vegas). The implication is that it may make sense to develop a rail safety and funding model that permits both heavy freight and other forms of rail service to co-exist by sharing track.

To coin a term, I've called this concept "rapid rail" to distinguish it from the prevailing notions of freight and HSR as animals that must be kept strictly separate at all times. The middle ground, which would be appropriate in many parts of the state, is to find a regulatory path to permit mixed traffic not as an exception but as the rule. The idea is to use existing freight tracks where they permit high speeds, such that only some sections require brand-new tracks. Achieving this will require the following:

  • new FRA rules spelling out the safety measures that must be implemented and enforced on network segments designated as "rapid rail". Sub-classes would permit all traffic, prohibit heavy freight, prohibit heavy and medium freight or, prohibit all freight for light cargo and passenger service only. Each segment on a rapid rail network would be mapped to one of these classes as appropriate.
  • public-private partnership (PPP) between the state and the freight operators plus Caltrain that owns the infrastructure. BART could join this partnership by installing gauge change stations at selected locations, plus retrofits to its rolling stock. The PPP would enjoy a 30-year monopoly franchise co-ordinating all planning and funding of infrastructure projects including both new alignment construction and, upgrades to and proper maintenance of the legacy portions. It would also be party to road, local transit, electric grid and urban densification planning. In terms of operations, it would act as the sole dispatcher of all heavy rail traffic, based on a timetable with adequate slots for heavy freight. It would be very counterproductive to favor passenger rail to such an extent that rail freight loses market share to trucking.

In addition, rapid rail segments would be gradually electrified at 25kV AC single phase @ 60Hz over a number of decades. Priority would go to long tunnels, sections affected by poor air quality and, high-volume commuter corridors suffering severe rush hour congestion. The latter creates pressure to invest in freeway upgrades, carries high opportunity costs due to loss of productive time and, high dependence on oil. Rail operators would need mandates and fiscal incentives to invest in two-mode locomotives or self-propelled trainsets to take immediate advantage of partial electrification of their routes. If it were to link to the standard gauge grid, BART would of course retain its third rail DC electrification.

A very important concept in this context is that HSR is not BART. While it is intended for true bullet trains, lightweight diesel or electric rolling stock could use the same tracks in the sections where the bullet trains don't run at more than 125mph anyhow. In particular, Caltrain, Metrolink, ACE and NCTD could - if they wanted to - choose to operate such purely regional HSR trains, using either fast DMU/EMU self-propelled rolling stock or even plain old locomotives such as this one:

To illustrate what I mean, I have created two maps showing in detail what this shift in regulatory philosophy and operational practice could help bring about. The first Rapid Rail map has three pages. Scroll down the icon list on the left if you want to see them all (btw, Google Maps has some bugs and sometimes fails to show everything at the first attempt). I encourage you to zoom in on details of interest to you and, to read notes I've attached to named lines and icons.

  • Page 1 shows an extensively built out rapid rail network for Northern California and the Central Valley. Highlights include:

    • HSR approach to SF via 101 freeway median (sacrifice traffic lanes but avoid cost of DTX tunnel)
    • loop track on 2nd floor of SF Transbay Terminal, buses in basement
    • dual tracks across Bay Bridge (sacrifice a traffic lane on each deck, not certain if bridge can take the load)
    • standard gauge rail-around-the-Bay. In the East Bay, this leverages an existing unused ROW adjacent to BART between the Union City and San Leandro stations. The section between Niles and hwy 262 would require a viaduct directly above UPRR and/or BART tracks. The nearby Hayward fault would complicate the civil engineering design.
    • a detour track past Oakland airport
    • a bypass route along hwy 4
    • a standard gauge intermodal station at Concord NWC (important intermodal with BART North Concord, eliminates need for eBART)
    • connection with CV towns via downtown Tracy
    • Amtrak San Joaquin moved to UPRR ROW
    • Altamont Pass connector (may eliminate need for BART extension to Livermore)
    • a new Capitol Corridor alignment via Vallejo
    • spur to Santa Rosa
    • spur Napa Valley
    • a new bridge and access segments to a loop track to Sacramento Airport terminals
    • alternative implementation of Sacramento HSR station to permit adjacent run-through tracks for rapid rail feeder trains (see Page 3 for details)
    • a Caltrain extension to Hollister
    • a new fast alignment from Gilroy to Monterey Cannery Row
    • a spur up to Santa Cruz Boardwalk along hwy 1.

    Note that the Benicia rail bridge and the western and northern approaches to it would become a dedicated freight corridor shared by UPRR and BNSF (cp. Alameda corridor in LA)

    For clarity, the complementary BART network is not shown.

  • Page 2 shows

    • a sped-up alignment for the Central Coast corridor routed inland around Vandenburg AFB. Check the Terrain view to see where this calls for the construction of new tunnels. Two-mode locomotives will be needed for the one near Solvang.
    • a rapid freight corridor between Bakersfield and Sylmar based on new tracks in the hwy 99 median plus a 48-mile base tunnel through the Grapevine. Electric traction would be mandatory in the tunnel, which heavy freight trains could easily traverse at speeds of well over 100mph in less than half an hour without expending a drop of diesel fuel. Similar long rail tunnels exist in Japan (Seikan), under the Channel between France and the UK and are in preparation in Switzerland (St. Gotthard), Austria/Italy (Brenner), France/Italy (Montblanc) and Spain/Morocco (Straits of Gibraltar). This one through the Grapevine would be the longest in the world, though and cross two active faults deep underground (cp. seismic risks of Seikan and Straits of Gibraltar tunnels).
    • dual standard gauge tracks across the Richmond-San Rafael bridge (at the expense of two precious traffic lanes, not certain if bridge can take the load)
    • a cargo distribution yard in the Concord NWC
    • a detour and freight/cargo access tracks to Castle Airport, if that is ever upgraded to commercial service for long distance passenger, dedicated cargo and heavy lift aircraft
    • SMART extensions to Tiburon and Sonoma town
    • a new HSR feeder network in the Visalia region
    • a new loop line in the Van Nuys area based on an old ROW
    • some railyards and other details

  • Page 3 shows details of my alternative concept for the Sacramento station. I would hate for that city to repeat LA's mistake of creating a terminus station without run-through tracks.

The second Rapid Rail map shows various options for getting around Southern California and over to Las Vegas. Please note in particular:

  • alternative #1: Anaheim - San Diego via existing tracks to San Juan Capistrano and new tracks in the I-5 median to Torrey Pines
  • alternative #2: Victorville - San Bernadino - Ontario - Anaheim - San Diego using upper floor tracks of Anaheim ARTIC that were intended for maglev to Las Vegas.

    Features new alignments on hwy 57 median and through Cajon Pass intended for passenger service for light/medium but not heavy freight.
  • loop to get from upper to lower floor tracks at ARTIC for direct LA US - San Diego service using this second option
  • alternative #3: Corona - San Diego (Balboa Park) via I-15 and hwy 163.
  • new Metrolink routes: LA US - Long Beach airport, LA US - Disneyworld - Anaheim loop. Long Beach airport could also be accessed by Orange County Metrolink routes that do not involve LA US. Electrified tunnels in both segments, two-mode locomotives required.
  • Victorville - Las Vegas as per Desert Xpress plans, i.e. privately funded and based on diesel trains at 125mph. Electrification optional but very highly recommended if alignment permits higher speeds that way. Consider mounting power distribution lines on the catenary masts to help defray the cost.
  • Mojave - Barstow connector to HSR starter line to permit access from Las Vegas to Palmdale airport, the Central Valley and Bay Area/Sacramento. Dual-mode trainsets using both HSR and Desert Xpress tracks must be capable of 186-220mph in electric mode, even if they are limited to 125mph when running on diesel.

UPDATE: I've added a third new map showing the option of a new Richmond - San Rafael rail bridge (h/t to David S.)

Of course, all of this goes well beyond the immediate objective of getting the HSR starter line built such that FRA permits the use of off-the-shelf non-compliant but proven trainsets on it. With its target top speed of 220mph, California HSR will immediately become one of the premier HSR services in the world.

My objective here was simply to point out future possibilities and especially, the conversion of traffic lanes on Bay Area bridges to standard gauge rail tracks. Getting to a judicious mix of car and rail cultures will require sacrificing some actual and potential freeway capacity. Also, a viable rail culture must absolutely meet the legitimate needs of the entire spectrum of rail operators, from passenger to heavy freight.


DoDo said...

You write that high speed rail is to short-haul flights as light rail is to buses. I beg to differ.

The different modes of local public transport don't just differ in technology and comfort: they also differ in ideal stopping distances and carrying capacity. In the latter, buses (and non-polluting trolley buses) are at the lower end, then comes light rail and suburban heavy rail, finally subways on the top.

This is an important point because the incorrectible opponents of public transport projects often resort to claims about cheaper alternatives -- where that 'alternative' might be too slow or low-capacity for the traffic corridor or city in question.

A better Argument #1 would be, IMHO, that high speed railways, conventional express and regional railways, suburban railways, metros/subways, light rail and (trolley) buses would all have their place in a coherent public transport system -- as it already exists in Europe and East Asia, or indeed even New York. The different modes enhance each other by giving passengers a better coverage of potential destinations without a car.

DoDo said...

Great work on the future rail network plan!

I will be critical about one thing: that 48-mile base tunnel.

Such tunnels are very expensive, and are projected at places of traffic bottlenecks -- and all of them are for every type of traffic. I.e. high-speed, too.

Seikan is to be rebuilt alongside the construction of the Hokkaido Shinkansen once the Tohoku line reaches it. Eurotunnel already carries the Eurostar trains, even if they slow down to 160 km/h in it. However, in the recentls opened Lötschberg Base Tunnel, top speed is 250 km/h. The three 50km+ alpine tunnels will be similar. (I note that the in-construction Gotthard Base Tunnel will be extended to c. 75km in a second stage, so your tunnel may face close competition :-))

So, North of Los Angeles, I think such a big investment would only be justified either if freight traffic is very heavy (to the tune of one train every five minutes) and pays for it, or if CHSRA runs its trains through it, too.

I looked at the alignment on Google's terrain map, for potential alternatives.

First, I notice that along your alignment, at the southern end, there is first a need for a base tunnel of only 4-5 miles in length. From then on, the route could follow Interstate 5, either bundled with the highway on the surface or in a cut-and-cover tunnel or a succession of both, to Castaic. But from then on, it's still a giant base tunnel of about 35 miles to Grapevine.

I think the cheapest option, with grades not worse than from Wheeler Ridge to Grapevine (c. 1.7%), would be this: an 18-mile tunnel from Grapevine to the Southeast to the c. 3000' high plateau, on the surface to Fairmont, a 10-mile tunnel down to the San Francisquito Canyon c. to the point where the valley bottom reaches 1800' (but with tunnel portal much higher), bridge across the valley, further descent on the Eastern valley side (it's not too winding) with some shorter tunnels and bridges, cut-and-cover tunnels in the built-up areas of Santa Clarita and Newhall, finally a 4 mile tunnel to San Fernando. The total tunnel length is about the same, but shorter and less deep tunnels are cheaper.

DoDo said...

Correction vs my first comment. If you meant that high speed rail is to short-haul flights as light rail is to bus rapid transit (BRT) -- hell yeah!

Rafael said...

@ DoDo -

1) we're in violent agreement that an effective transit system needs complementary and interconnected services at multiple speed and distance scales. You're right about light rail vs. buses, that should have been trolleys vs. buses (or light rail vs. BRT, as you pointed out). I've made the correctionn.

2) for reference, the cost estimates for the Brenner base tunnel (56km = 35mi) have escalated alarmingly from EUR 4.5 billion to EUR 9 billion. Many are now arguing that the project should proceed only if there are ironclad guarantees that trucks really will be forced off the extremely congested Inntal highway. Note that the tunnel would also require fairly massive investments in access routes to both ends.

In California, that would be less of an issue, since rails almost to the southern entrance already exist. The northern approach is an almost straight line to Bakersfield across level terrain. Still, base tunnel projects are never cheap, so they had better be worth it.

Austria already operates trains that ferry trucks through the Alps along the existing rail line. The idea is that trucks can keep moving while drivers take their mandatory breaks, but the service is fairly expensive. EuroTunnel operates car-and-truck trains through the Channel Tunnel. It works pretty much like a really fast ferry.

I'd have no problem with letting bullet trains use a shortcut throught the Grapevine, that would certainly greatly increase the number of trains actually using the tunnel. A heavy rail shuttle between LAUS and Long Beach airport (see map #2) could be implemented at moderate cost to compensate for the loss of much more remote Palmdale as a relief airport for LAX. An added benefit would be reduced population growth in the High Desert and hence, reduced pressure to increase the flow of scarce water to that area.

However, there are some issues:

- in the event of a major earthquake, a train might well derail. Even if no-one is severely injured, the logistics involved in rescuing hundreds of passengers 20 miles from the tunnel entrances are nothing if not daunting - even with a third tube. Fire poses a separate serious risk in all long tunnels, irrespective of geology.

- any significant elevation gain at all will force US-style heavy freight trains to slow down substantially, creating a traffic jam for passenger trains. The beauty of a single tunnel is that the elevation difference between Grapevine and Sylmar is just 200 feet over 48 miles. That means heavy freight trains could thunder through at 100+ mph. The section along the hwy 99 median would allow similarly high speeds.

The only reason FRA limits heavy freight trains to 79mph is because private operators aren't prepared to invest in the signaling needed to make higher speeds possible. On straight, level alignments, the locomotives could manage quite a bit more, though the already massive kinetic energy would of course grow with the square of velocity. Rapid freight trains have very long emergency brake distances, so cruise speeds might have to be optimized for maximum throughput rather than minimum transit time.

- FRA would surely require unprecedented safeguards against train-on-train collisions in a high speed corridor shared by 15,000 ton FRA-compliant freight behemoths and non-compliant bullet trains. At this point, they'd probably laugh you at of the room at the mere suggestion.

David S said...

I don't know about bridge load issues (although the entire bridge was recently retrofitted for seismic safety - on time and under budget, no less!) but the Richmond - San Rafael bridge decks are both 3 lanes which, but only 2 of which are allowed for general use. The third lane is a shoulder/emergency lane.

At one point, I believe, the bridge had 3 lanes, but one lane was removed to add a water pipe. They never returned to 3x3 because the marin approach only has 2 lanes in either direction, which would just shift the bottle neck around.

The bridge also doesn't regularly see massive slowdowns on the span itself, rather, the slowdowns happen at the 580/101 interchange(s) and at the toll plaza.

IMO, a BART or SMART extension would be a worthy enough cause to sacrifice the emergency lane for. Marin and Sonoma are drastically under-served by public transit services.

Great post!

Rafael said...

@ David S. -

the emergency lanes on the Richmond - San Rafael road bridge were used for traffic after the Loma Prieta earthquake knocked out the Bay Bridge for a while. That was a strictly temporary arrangement, though. Caltrans insists that an emergency lane is needed on that bridge to maintain safety.

The water pipe was originally installed to help Marin county cope with severe drought spell. If indeed it was left in place, that's because the next drought is sure to come. Moving it to run under the bridge would free up one traffic lane, possibly enabling at least a single standard gauge track across, without having to sacrifice any traffic lanes.

The alternative would be to build a new at-grade dual track rail bridge via the larger of the two small islands located just north of the road bridge. I believe there's just a B&B there at the moment. The eastern span would get a bascule section to keep the shipping lane open. This would make sense only if both anticipated rail and anticipated shipping traffic are light enough to avoid an excessive number of bascule operations.

The eastern approach already has a rail line, though it runs through the Chevron refinery and is a little twisty in places. Nothing a short tunnel couldn't fix, though.

On the Marin side, the approach would run south, parallel to Shoreline Pkwy. Threading a needle through the narrowest portion of the built-up area, it would pass under I-580 and veer north through some shrubbery to join up with the SMART ROW.

Both the San Rafael SMART and Richmond Amtrak/BART stations would be served.

For clarity, I've created a Richmond-San Rafael rail bridge map.

Note that tranportation planners may insist on a bridge design strong enough to support even heavy freight trains, unless the old ROW through the sloughs between

Rafael said...

... between Novato and American Canyon is restored. NCRA plans to do just that for NWP freight trains bearing ore.

What they neglected to mention is that NWP intends to run its mile-long trains at night to minimize conflicts with road traffic and SMART operations. The NIMBYs are not happy campers about this.

Spokker said...

Rafael, what do you do for a living and what are your career prospects?

It appears that you should be working on California's rail network full-time instead of the clowns that are on some of these public agencies.

DoDo said...

@ Rafael -

I'm not read-up on the BBT cost estimates at the moment, but venture a guess that similar reasons are at play as have been for the GBT and the French-Italian project, which I know better. There, the main cost-boosters were: increased safety requirements, the financing effect of delays and re-tenderings, and unexpected geology. Of these factors, I think only the second can be avoided. (As for access routes, while not inexpensive, their cost is dwarfed by that of the main tunnels.)

I see I was unclear about the dual use argument. What I meant was that the cost-benefit calculations for the GBT or Lyon-Torino include co-use by high-speed transit, while CHSRA will have its own route (with higher grades being much less of a problem for their trains), thus this option is bust. But, good point about the possibility of piggybacking (or, using the nice German name for it, Rollende Landstraße = "rolling country road").

On grades. For ventillation and water conduction reasons, most large rail tunnels have a (small) grade towards a midpoint, for example, the climb from the South portal in the GBT is about 240 m (c. 800 feet), with a 0.8% grade. So I assumed maximum grade as before Grapevine, though now I see that's a bit much.

On high-speed freight -- a bit of a diversion for my curiosity. You say US railfreight could go much faster than 79 mph, and you're not the first I hear this from. Now, I'm not too familiar with US rail freight car bogies, and hence my question: how do they master higher speeds? Here in Europe, just 160 km/h (100 mph) requires a good two-level sprung/damped suspension (side bearings, high critical speed of hunting), while most railfreight rides cars with robust bogies with one-level or 'one-and-half level' sprung suspension (i.e. elastic bogie pivot) for 120 km/h (75 mph), even LEILA (

ry said...

Did that locomotive actually "play" a major scale as it accelerated to depart the station?

Rafael said...

@ DoDo -

I'm not intimately familiar with the details of suspension systems for heavy freight rolling stock. However, I am sure that freight operators would be willing to invest in some rolling stock that would permit the sustained speeds necessary to make such a tunnel project pencil out.

As for the gradient, keep in mind that US freight trains are 2-3 as massive as their counterparts in Europe.

It would be interesting to learn how much the HSR detour via Palmdale is expected to costs, relative to such a base tunnel. I expect the latter would be substantially more expensive, but IFF the FRA figures out how a base tunnel could be shared by HSR and heavy freight, it would also be more valuable to the California economy as a whole.

The impact of shifting freight from road to (electrified) rail on air quality in the LA basin alone would be significant and valuable.

@ ry -

it is a C scale, performed by the power electronics that represent the solid-state transmission of electric traction systems.

timote said...


Even with the mixed use, wouldn't they want to pick up Palmdale for commuter reasons? It's like the Altamont discussion but more justifiable due to money...

Rafael said...

@ timote -

afaik, the primary drivers in favor of Tehachapi Pass were geology (major fault crossings at grade) and the desire to connect relief airports directly to the HSR line.

The chosen alignment is based on gradients of up to 3.5%, far too high for heavy freight trains to negotiate. They're already down to bicycle speeds at 2.2%.

Also, the high desert is at around 3200 feet above MSL, whereas the LA basin (at Sylmar) is at around 1400 and the Central Valley (at Grapevine) at around 1200. So while freight trains can take the route via Palmdale, they have to use a whole lot of fuel and brake linings to negotiated a 2000ft difference in elevation in transit between two points whose elevations differ by 1/10th of that.

timote said...

Gotcha, your argument is that the choice of Palmdale pretty much excludes freight mixed-use, that with the proposed route it will be permanently passenger-only.

nikko pigman said...

Wow raphael, this is really great work! I disagree with you on some points and I would say others are arguable but this is all in all well conceived. I plan on going into public transportation and I may keep this idea in mind in the future.

I have the same question as Spokker, what is your profession or if you have yet to establish one what are you planning to do?

Rafael said...

@ spokker, nikko pigman -

I'm a mechanical engineer by trade, but my interests range a little wider. This HSR thing is just a hobby.

Rafael said...

@ timote -

light/medium freight trains can and do storm up electrified alignments with gradients as high as 5.5% in e.g. Switzerland, but that's because they are competing against trucking at intermediate distances. European freight trains are shorter than their US counterparts and, they don't stack shipping containers on top of one another. Instead, the focus is on things like car delivery

However, US-style heavy freight trains could not negotiate the 3.5% gradients CHSRA will use for its alignment. Besides, high speed means just that and therefore, lightweight UIC-compliant trainsets with proven track records. FRA will not allow those to share track with FRA-compliant rolling stock.

What is possible and according to the new business plan, intended, is the operation of light cargo versions of the same trainsets used for passengers. This refers to goods with a high value per unit mass, e.g. mail, packages, perishable foodstuffs, fresh flowers, light manufacturing parts (e.g. computer chips), even air cargo containers.

There is a precedent in the yellow La Poste TGV trains in France. Note the double trainset duplex (bi-level) passenger TGV traveling in the other directions - it has around 1100 seats. SNCF reports an average seat capacity utilization on its TGVs.

It would be possible to operate cargo trains at night, albeit at more moderate speeds, if the problem of rail-wheel and aerodynamic noise can be dealt with. That's something the Japanese and the Spaniards (Talgo) are very good at. The tracks are there 24/7, so why not use them.

Another option is to attach self-propelled cargo trainsets to single trainset passenger trains to avoid impacting headways between passenger trains. Coupling of HSR trainsets at stations is a fast operation.

However, to enable high speed cargo operations, CHSRA would have to include sidings and/or spurs off its main line. This is trickier than it sounds because those cargo trains also must not share track with FRA-compliant freight or passenger trains. That means cargo transshipment centers have to be located very close to the HSR line and integrated into plans just like passenger stations. Afaik, that work hasn't been done yet and it's not really trivial to do as an afterthought because of the impact on passenger operations.

BruceMcF said...

Great post ... I've quoted two paragraphs and linked back in Dear Joe, I want an Electric Train for Christmas

BruceMcF said...

@ DoDo "On high-speed freight -- a bit of a diversion for my curiosity. You say US railfreight could go much faster than 79 mph, and you're not the first I hear this from. Now, I'm not too familiar with US rail freight car bogies, and hence my question: how do they master higher speeds?"

110mph is is substantially higher speed than 79mph, especially if sustained (as is not uncommon across the plains). And that extra speed is important for cracking schedule sensitive and time sensitive freight market share, in recovering the time required for intermodal operations and marshalling as opposed to dock to dock truck freight.

If on the same plan as the various land bridge proposals in Australia, if the support infrastructure can take the axle loading for 79mph double-stack, it can take somewhere around 70% the axle loading for 110mph single-stack.

So while I have no idea what the detailed answer is, assuming that similar stresses are in place between the track and the foundation as between the track and the car body, then the simple answer is that to go faster than 79mph, you don't double stack and the maximum axle loading is scaled down.

Rafael said...

@ BruceMcF -

thx for the shout-out over at daily Kos etc.

Wrt rapid freight, all I said was that the locomotives can handle higher speeds. Double-stacking may or may not be possible at those higher speeds, that's something the engineers designing the track bed would have to figure out.

In any case, even at 79mph, it would take a heavy freight train just over half an hour to traverse a base tunnel through the Grapevine - a fraction of the time it takes to get through Tehachapi Pass. Any passenger trains also using the tunnel would have a chance to overtake freight trains between the northern tunnel entrance and the hwy 99 median. Indeed, that portion of the route could be next to hwy 99 or even up the hwy 184 corridor so additional bypass sections can be built. That would also allow the shortcut to connect back to existing lines east of Bakersfield, so passenger trains could stop at Thruxton Ave.

Those are details, though. The thrust of my argument was that passenger service isn't the only thing that could be - perhaps should be - electrified and sped up.

Anonymous said...

Any rail transit system for the North Bay should utilize the Golden Gate Bridge not the Richmond San Rafael Bridge. Golden Gate Transit's #80 trunk bus route would be replaced by an electric standard gauge light rail system the would access San Francisco via a shared subway with SF Muni.

The GG Bridge is underutilized, especially in light of its high maintenance costs. It cries out to be reconfigured to handle light rail. The problem is not engineering but political - the Bridge District has always wanted a second deck for lucrative toll-paying automobiles.

But all rail improvements in the Bay Area are now probably moot since BART has shoved thru its San Jose extension. With its political connections BART will hoover up every penny of the infrastructure monies.